The objectives of the proposed work are to develop theoretical and computational methods that can be used to interpret the behavior of biological molecules, and to apply these methods to help explain the activity of a variety of specific systems of biological interest. A major focus in the first area will be the combination of statistical mechanical and molecular dynamics simulation techniques to produce computational methods for predicting free energy changes associated with the simultaneous variation of molecular composition and geometry. Such methods should be very useful for predicting how ligand-binding affinities and chemical reaction rates change with changes in molecular composition, for example, and the proposed work includes developmental studies aimed at making both these kinds of predictions practical. The new computational methods will be used to provide accurate and detailed descriptions of the nature of molecular recognition in trypsin and its mutants, in L-arabinose-binding protein, in antibodies, and in picornaviruses. Also, methods for simulating activated processes will be used to study solvent effects on the dynamics of conformational transitions and ligand binding in different proteins. This work should provide useful new tools for designing drug, enzymes and other molecules for practical applications, and the deepening of understanding of the activity of biological molecules. This proposal was jointly review by the Biophysics Program, the Physical Chemistry Program and the Biological Instrumentation Program and all three programs will participate in the funding of this research project.
